Perceptron. In order to train a perceptron, we will simplify the family car problem. Specifically, we will remove the examples for Price=20000, reduce the magnitude of the values for Price and EnginePower, and change the class values to be 0 (no) or 1 (yes). So, there are now 10 training examples as follows.

We will train the perceptron below until it correctly classifies all 10 of these training examples. Training is accomplished by updating the weights according to the perceptron learning rule (equation 18.7 in the textbook). Assume the initial weights are all equal to 1.0, and a = 0.5. Consider each example in the order presented in the table above and show the weight updates for each incorrectly-classified example. Continue until the perceptron correctly classifies all the training examples, and show the final perceptron weights.

Price EnginePower CupholdersFamilyCar 10000 100 200 200 300 300 100 100 200 300 300 100 100 200 200 300 no 10000 10000 10000 20000 no yes yes yes yes no yes no no 20000 30000 no 30000 no Price EnginePower CupholdersFamilyCar 10000 100 200 200 300 300 100 100 200 300 300 100 100 200 200 300 no 10000 10000 10000 20000 no yes yes yes yes no yes no no 20000 30000 no 30000 no